Gas Pipelines are usually designed to operate in steady state, but in practice most pipelines deliver transient loads. Transient loads degrade total delivery capacity and lose revenue. This paper uses simple examples to quantify such losses and shows how to keep from losing more in practice. The challenge to pipeline operators is to minimize lost capacity by choosing opportunities wisely and exercising skillful control. Look-forward hydraulic simulation has long been used to evaluate spot transportation opportunities and their impact on prior delivery commitments. Gas Control tests control strategies with simulation using a “What-if Study”: If a load appears, “What if we respond this way?” The control selection is often driven by the number of alternate strategies that can be analyzed in the time available for a decision. Thus, while the selected strategy might be reasonable, it may not be optimal for the pipeline company, just the best they had time to try. Through studying example events, this paper describes a completely new approach to What-if Studies. It introduces the use of real-time transient optimization technology to allow the engineer, operator, or even the marketer to get answers for the completely different “What-if” question, namely “What control strategy is best if the network must respond to a specific event?” This replaces manual-iterative look-ahead simulations by mathematical optimization. For each possible load event this cutting-edge technology presents the highest-profit strategy, if one is feasible. Moreover, if delivering the load is infeasible it can determine the alternate strategy that makes the most profit, such as which loads to shed, when, and by how much.
Latent likelihood ratio tests for assessing spatial kernels in epidemic models
